[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-users] Driver domain (PV domain) crashes with PCI-passthrough of RAID controller


  • To: xen-users@xxxxxxxxxxxxx
  • From: Alex Wakizashi <alex@xxxxxxxxxxxxxx>
  • Date: Fri, 7 Apr 2017 02:51:18 +0300
  • Delivery-date: Thu, 06 Apr 2017 23:52:28 +0000
  • List-id: Xen user discussion <xen-users.lists.xen.org>

Hello, all,

Have tried to use driver domain, but have encountered following
problem - when using PV domain (or PVH, same problem) with PCI
passthrough on Adaptec RAID controller, aacraid driver hangs.
Destroying domain causes host(!!!) restart.
Same happens if I trying to detach device from domain.

It's working perfectly with same VM image, started in HVM mode, but
AFAIK(please, correct, if I'm wrong) it's not possible to use
driver-domains in HVM mode.

Have tried both adding "e820host=1" and disabling MSI - does not help.
Have played with CPU affinity - so VM are working on same NUMA node,
RAID controller attached to, no difference.
(Previously, that helped to start PCI-passthrough for NVidia Quadro -
as BIOS set the routing of VGA IO to BMC VGA adapter, and GPU VBIOS
just hangs, when GPU was on 2-nd NUMA node's PCIe bus - due to crazy
IO routing).
Unfortunately, it's not possible to install RAID controller to CPU#0
(NUMA node #0) PCIe bus - but looking to the crash message, it seems
to be related...

XEN is stable-4.8, kernel is vanilla 4.9.20, compiled "debian way",
systems (both host and guest) are Debian 8 (Jessie).

Any ideas/assumptions how to solve it?

Driver domains seems to be quite important feature, sad that it's not
working well...

Details:

When starting VM, following messages appears in XEN's dmesg:

[  396.541483] pciback 0000:83:00.0: enabling device (0000 -> 0003)
[  396.541565] xen: registering gsi 64 triggering 0 polarity 1
[  396.541576] Already setup the GSI :64
[  396.541927] xen: registering gsi 64 triggering 0 polarity 1
[  396.541933] Already setup the GSI :64
[  396.542178] xen: registering gsi 64 triggering 0 polarity 1
[  396.542183] Already setup the GSI :64
[  396.542429] xen: registering gsi 64 triggering 0 polarity 1
[  396.542434] Already setup the GSI :64
[  396.542687] xen: registering gsi 64 triggering 0 polarity 1
[  396.542693] Already setup the GSI :64
[  396.557114] xen-blkback: backend/vbd/4/768: using 2 queues,
protocol 1 (x86_64-abi) persistent grants
[  396.592680] vif vif-4-0 vif4.0: Guest Rx ready
[  396.592758] IPv6: ADDRCONF(NETDEV_CHANGE): vif4.0: link becomes ready
[  396.592803] xenbr0: port 2(vif4.0) entered blocking state
[  396.592808] xenbr0: port 2(vif4.0) entered forwarding state
(XEN) [2017-05-08 02:14:39.261] traps.c:3715: GPF (0000):
ffff82d0801b11cc -> ffff82d080260507
(XEN) [2017-05-08 02:14:39.261] traps.c:3715: GPF (0000):
ffff82d0801b11cc -> ffff82d080260507
(XEN) [2017-05-08 02:15:21.320] d4: Forcing read-only access to MFN fbc02
(XEN) [2017-05-08 02:15:21.320] d4: Forcing read-only access to MFN fbc03
(XEN) [2017-05-08 02:15:21.330] d4: Forcing read-only access to MFN fbc02
(XEN) [2017-05-08 02:15:21.330] d0: Forcing read-only access to MFN fbc02


Worst thing is, when trying to destroy domain, whole host (!!) just
rebooting with message:

(XEN) [2017-05-08 02:16:08.673] irq.c:2140: dom4: forcing unbind of pirq 103
(XEN) [2017-05-08 02:16:08.673] irq.c:2140: dom4: forcing unbind of pirq 102
(XEN) [2017-05-08 02:16:08.674] Assertion 'pos' failed at msi.c:1293
(XEN) [2017-05-08 02:16:08.674] Debugging connection not set up.
(XEN) [2017-05-08 02:16:08.674] ----[ Xen-4.8.1-pre  x86_64  debug=y
Not tainted ]----
(XEN) [2017-05-08 02:16:08.674] CPU:    5
(XEN) [2017-05-08 02:16:08.674] RIP:    e008:[<ffff82d08018734f>]
pci_msi_conf_write_intercept+0x7d/0x20d
(XEN) [2017-05-08 02:16:08.674] RFLAGS: 0000000000010246   CONTEXT:
hypervisor (d0v5)
(XEN) [2017-05-08 02:16:08.674] rax: 0000000000000000   rbx:
ffff83083953cf10   rcx: 0000000000000001
(XEN) [2017-05-08 02:16:08.674] rdx: ffff83083955ffff   rsi:
0000000000000293   rdi: ffff82d080340180
(XEN) [2017-05-08 02:16:08.674] rbp: ffff83083955fd58   rsp:
ffff83083955fd18   r8:  0000000000000000
(XEN) [2017-05-08 02:16:08.674] r9:  ffff82d0803642c8   r10:
0000000000000000   r11: 0000000000001c06
(XEN) [2017-05-08 02:16:08.674] r12: 00000000000000b2   r13:
ffff83083955fe04   r14: 0000000000000000
(XEN) [2017-05-08 02:16:08.674] r15: 0000000000000083   cr0:
000000008005003b   cr4: 00000000001526e0
(XEN) [2017-05-08 02:16:08.674] cr3: 0000000835c07000   cr2: ffff88066b5889a0
(XEN) [2017-05-08 02:16:08.674] ds: 0000   es: 0000   fs: 0000   gs:
0000   ss: 0000   cs: e008
(XEN) [2017-05-08 02:16:08.674] Xen code around <ffff82d08018734f>
(pci_msi_conf_write_intercept+0x7d/0x20d):
(XEN) [2017-05-08 02:16:08.674]  6b 8c fc ff 85 c0 75 02 <0f> 0b 44 39
e0 77 4f 8d 50 0c 41 39 d4 73 47 83
(XEN) [2017-05-08 02:16:08.674] Xen stack trace from rsp=ffff83083955fd18:
(XEN) [2017-05-08 02:16:08.674]    ffff83083955fd58 0000000200002bd7
ffff83083955fd48 00000000000000b2
(XEN) [2017-05-08 02:16:08.674]    0000000000000000 0000000000008300
0000000000000002 0000000000000000
(XEN) [2017-05-08 02:16:08.674]    ffff83083955fda8 ffff82d0801a28d8
00004c0897255878 ffff83083955fe04
(XEN) [2017-05-08 02:16:08.674]    000000ffd5554239 ffff83083ff0b000
ffff83083955fe04 0000000000000002
(XEN) [2017-05-08 02:16:08.674]    00000000000000b2 0000000000008300
ffff83083955fde8 ffff82d0801afe7f
(XEN) [2017-05-08 02:16:08.674]    ffff82d08017a294 0000000000000cfe
0000000000000002 0000000000000002
(XEN) [2017-05-08 02:16:08.674]    ffff83083ff0b000 ffff83083955fe04
ffff83083955fe38 ffff82d0801b333e
(XEN) [2017-05-08 02:16:08.674]    ffff83083955fe28 00007fff80125db1
0000000000000246 ffff83083955ff18
(XEN) [2017-05-08 02:16:08.674]    ffff83007daf4000 0000000000000cfe
ffff8200400142c0 ffff83083ff0b000
(XEN) [2017-05-08 02:16:08.674]    ffff83083955ff08 ffff82d0801b5c8d
ffff830839564140 000000023955ff18
(XEN) [2017-05-08 02:16:08.674]    ffff83007daf4000 ffff83083ff0b000
ffff830800000000 0000000800000001
(XEN) [2017-05-08 02:16:08.674]    02ff82d0801d4100 0000000000000000
ffff82d0bfffa2c0 ffffffff00000002
(XEN) [2017-05-08 02:16:08.674]    ffffffff81685184 deadbeef00000008
ffff83083ff0b000 ffff830800a0fb00
(XEN) [2017-05-08 02:16:08.674]    0000000000000000 ffffffffffffffff
ffff83007daf4000 00000000ffffffef
(XEN) [2017-05-08 02:16:08.674]    ffff83083952f000 ffff83007daf4000
0000000000007fff 0000000000000002
(XEN) [2017-05-08 02:16:08.674]    0000000000000000 00000000000000b2
00007cf7c6aa00c7 ffff82d08025d5c8
(XEN) [2017-05-08 02:16:08.674]    00000000000000b2 0000000000000000
0000000000000002 0000000000007fff
(XEN) [2017-05-08 02:16:08.674]    0000000000000002 0000000000830000
0000000000000000 0000000000000000
(XEN) [2017-05-08 02:16:08.674]    0000000000007fff 0000000000000002
0000000000007fff 0000000000000000
(XEN) [2017-05-08 02:16:08.674]    0000000000000cfe ffffffff816800b0
ffffffff81f7d030 0000000d00000000
(XEN) [2017-05-08 02:16:08.674] Xen call trace:
(XEN) [2017-05-08 02:16:08.674]    [<ffff82d08018734f>]
pci_msi_conf_write_intercept+0x7d/0x20d
(XEN) [2017-05-08 02:16:08.674]    [<ffff82d0801a28d8>]
pci_conf_write_intercept+0x99/0xb5
(XEN) [2017-05-08 02:16:08.674]    [<ffff82d0801afe7f>]
traps.c#pci_cfg_ok+0x10f/0x17d
(XEN) [2017-05-08 02:16:08.674]    [<ffff82d0801b333e>]
guest_io_write+0x185/0x1fb
(XEN) [2017-05-08 02:16:08.674]    [<ffff82d0801b5c8d>]
do_general_protection+0x195b/0x2800
(XEN) [2017-05-08 02:16:08.674]    [<ffff82d08025d5c8>]
entry.o#handle_exception_saved+0x66/0xa4
(XEN) [2017-05-08 02:16:08.674]
(XEN) [2017-05-08 02:16:09.063]
(XEN) [2017-05-08 02:16:09.066] ****************************************
(XEN) [2017-05-08 02:16:09.072] Panic on CPU 5:
(XEN) [2017-05-08 02:16:09.076] Assertion 'pos' failed at msi.c:1293
(XEN) [2017-05-08 02:16:09.082] ****************************************
(XEN) [2017-05-08 02:16:09.089]
(XEN) [2017-05-08 02:16:09.092] Reboot in five seconds...
(XEN) [2017-05-08 02:16:09.097] Debugging connection not set up.
(XEN) [2017-05-08 02:16:09.103] [VT-D]iommu.c:1084: Set iommu
interrupt affinity error!
(XEN) [2017-05-08 02:16:14.113] Resetting with ACPI MEMORY or I/O RESET_REG.


Guest's logs:

(d4) [2017-05-08 02:14:35.400] [    0.000000] Linux version
4.9.0-ogun1-amd64 (os@xxxxxxxx) (gcc version 4.9.2 (Debian
4.9.2-10+ogun3+cert4) ) #1 SMP Debian 4
(d4) [2017-05-08 02:14:35.400] .9.20-1+ogun2 (2017-04-04)
(d4) [2017-05-08 02:14:35.400] [    0.000000] Command line: debug
earlyprintk=xen root=/dev/xvda2 iommu=soft
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/fpu: Supporting
XSAVE feature 0x001: 'x87 floating point registers'
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/fpu: Supporting
XSAVE feature 0x002: 'SSE registers'
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/fpu: Supporting
XSAVE feature 0x004: 'AVX registers'
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/fpu:
xstate_offset[2]:  576, xstate_sizes[2]:  256
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/fpu: Enabled xstate
features 0x7, context size is 832 bytes, using 'standard' format.
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/fpu: Using 'eager'
FPU context switches.
(d4) [2017-05-08 02:14:35.400] [    0.000000] ACPI in unprivileged
domain disabled
(d4) [2017-05-08 02:14:35.400] [    0.000000] Released 0 page(s)
(d4) [2017-05-08 02:14:35.400] [    0.000000] e820: BIOS-provided
physical RAM map:
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x0000000000000000-0x000000000009ffff] usable
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000000a0000-0x00000000000fffff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x0000000000100000-0x000000007df1dfff] usable
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x000000007df1e000-0x000000007e0d3fff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x000000007e0d4000-0x000000007e2dbfff] ACPI NVS
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x000000007e2dc000-0x000000007f357fff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x000000007f358000-0x000000007f7fffff] ACPI NVS
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x0000000080000000-0x000000008fffffff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000cfffc000-0x00000000cfffcfff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000fbffe000-0x00000000fbffefff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000fec00000-0x00000000fec01fff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000fec40000-0x00000000fec40fff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000fed1c000-0x00000000fed3ffff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000fee00000-0x00000000feefffff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x00000000ff000000-0x00000000ffffffff] reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] Xen: [mem
0x0000000100000000-0x00000001420e1fff] usable
(d4) [2017-05-08 02:14:35.400] [    0.000000] bootconsole [xenboot0] enabled
(d4) [2017-05-08 02:14:35.400] [    0.000000] NX (Execute Disable)
protection: active
(d4) [2017-05-08 02:14:35.400] [    0.000000] DMI not present or invalid.
(d4) [2017-05-08 02:14:35.400] [    0.000000] Hypervisor detected: Xen
(d4) [2017-05-08 02:14:35.400] [    0.000000] e820: update [mem
0x00000000-0x00000fff] usable ==> reserved
(d4) [2017-05-08 02:14:35.400] [    0.000000] e820: remove [mem
0x000a0000-0x000fffff] usable
(d4) [2017-05-08 02:14:35.400] [    0.000000] e820: last_pfn =
0x1420e2 max_arch_pfn = 0x400000000
(d4) [2017-05-08 02:14:35.400] [    0.000000] MTRR: Disabled
(d4) [2017-05-08 02:14:35.400] [    0.000000] x86/PAT: MTRRs disabled,
skipping PAT initialization too.
(d4) [2017-05-08 02:14:35.401] [    0.000000] x86/PAT: Configuration
[0-7]: WB  WT  UC- UC  WC  WP  UC  UC
(d4) [2017-05-08 02:14:35.401] [    0.000000] e820: last_pfn = 0x7df1e
max_arch_pfn = 0x400000000
(d4) [2017-05-08 02:14:35.401] [    0.000000] Scanning 1 areas for low
memory corruption
(d4) [2017-05-08 02:14:35.401] [    0.000000] Base memory trampoline
at [ffff88000009a000] 9a000 size 24576
(d4) [2017-05-08 02:14:35.401] [    0.000000] BRK [0x01f8d000,
0x01f8dfff] PGTABLE
(d4) [2017-05-08 02:14:35.401] [    0.000000] BRK [0x01f8e000,
0x01f8efff] PGTABLE
(d4) [2017-05-08 02:14:35.402] [    0.000000] BRK [0x01f8f000,
0x01f8ffff] PGTABLE
(d4) [2017-05-08 02:14:35.402] [    0.000000] BRK [0x01f90000,
0x01f90fff] PGTABLE
(d4) [2017-05-08 02:14:35.403] [    0.000000] BRK [0x01f91000,
0x01f91fff] PGTABLE
(d4) [2017-05-08 02:14:35.403] [    0.000000] BRK [0x01f92000,
0x01f92fff] PGTABLE
(d4) [2017-05-08 02:14:35.404] [    0.000000] BRK [0x01f93000,
0x01f93fff] PGTABLE
(d4) [2017-05-08 02:14:35.404] [    0.000000] BRK [0x01f94000,
0x01f94fff] PGTABLE
(d4) [2017-05-08 02:14:35.404] [    0.000000] BRK [0x01f95000,
0x01f95fff] PGTABLE
(d4) [2017-05-08 02:14:35.405] [    0.000000] BRK [0x01f96000,
0x01f96fff] PGTABLE
(d4) [2017-05-08 02:14:35.405] [    0.000000] BRK [0x01f97000,
0x01f97fff] PGTABLE
(d4) [2017-05-08 02:14:35.406] [    0.000000] BRK [0x01f98000,
0x01f98fff] PGTABLE
(d4) [2017-05-08 02:14:35.831] [    0.000000] NUMA turned off
(d4) [2017-05-08 02:14:35.831] [    0.000000] Faking a node at [mem
0x0000000000000000-0x00000001420e1fff]
(d4) [2017-05-08 02:14:35.831] [    0.000000] NODE_DATA(0) allocated
[mem 0x7d933000-0x7d937fff]
(d4) [2017-05-08 02:14:35.852] [    0.000000] Zone ranges:
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA      [mem
0x0000000000001000-0x0000000000ffffff]
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA32    [mem
0x0000000001000000-0x00000000ffffffff]
(d4) [2017-05-08 02:14:35.852] [    0.000000]   Normal   [mem
0x0000000100000000-0x00000001420e1fff]
(d4) [2017-05-08 02:14:35.852] [    0.000000]   Device   empty
(d4) [2017-05-08 02:14:35.852] [    0.000000] Movable zone start for each node
(d4) [2017-05-08 02:14:35.852] [    0.000000] Early memory node ranges
(d4) [2017-05-08 02:14:35.852] [    0.000000]   node   0: [mem
0x0000000000001000-0x000000000009ffff]
(d4) [2017-05-08 02:14:35.852] [    0.000000]   node   0: [mem
0x0000000000100000-0x000000007df1dfff]
(d4) [2017-05-08 02:14:35.852] [    0.000000]   node   0: [mem
0x0000000100000000-0x00000001420e1fff]
(d4) [2017-05-08 02:14:35.852] [    0.000000] Initmem setup node 0
[mem 0x0000000000001000-0x00000001420e1fff]
(d4) [2017-05-08 02:14:35.852] [    0.000000] On node 0 totalpages: 786335
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA zone: 64 pages
used for memmap
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA zone: 21 pages reserved
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA zone: 3999 pages,
LIFO batch:0
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA32 zone: 7997 pages
used for memmap
(d4) [2017-05-08 02:14:35.852] [    0.000000]   DMA32 zone: 511774
pages, LIFO batch:31
(d4) [2017-05-08 02:14:35.862] [    0.000000]   Normal zone: 4228
pages used for memmap
(d4) [2017-05-08 02:14:35.862] [    0.000000]   Normal zone: 270562
pages, LIFO batch:31
(d4) [2017-05-08 02:14:35.865] [    0.000000] p2m virtual area at
ffffc90000000000, size is 40000000
(d4) [2017-05-08 02:14:36.421] [    0.000000] Remapped 270562 page(s)
(d4) [2017-05-08 02:14:36.421] [    0.000000] smpboot: Allowing 2
CPUs, 0 hotplug CPUs
(d4) [2017-05-08 02:14:36.421] [    0.000000] e820: [mem
0x90000000-0xcfffbfff] available for PCI devices
(d4) [2017-05-08 02:14:36.421] [    0.000000] Booting paravirtualized
kernel on Xen
(d4) [2017-05-08 02:14:36.421] [    0.000000] Xen version: 4.8.1-pre
(preserve-AD)
(d4) [2017-05-08 02:14:36.421] [    0.000000] clocksource:
refined-jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns:
7645519600211568 ns
(d4) [2017-05-08 02:14:36.421] [    0.000000] setup_percpu:
NR_CPUS:512 nr_cpumask_bits:512 nr_cpu_ids:2 nr_node_ids:1
(d4) [2017-05-08 02:14:36.422] [    0.000000] percpu: Embedded 36
pages/cpu @ffff880141e00000 s108824 r8192 d30440 u1048576
(d4) [2017-05-08 02:14:36.422] [    0.000000] pcpu-alloc: s108824
r8192 d30440 u1048576 alloc=1*2097152
(d4) [2017-05-08 02:14:36.422] [    0.000000] pcpu-alloc: [0] 0 1
(d4) [2017-05-08 02:14:36.422] [    0.000000] xen: PV spinlocks enabled
(d4) [2017-05-08 02:14:36.422] [    0.000000] PV qspinlock hash table
entries: 256 (order: 0, 4096 bytes)
(d4) [2017-05-08 02:14:36.422] [    0.000000] Built 1 zonelists in
Node order, mobility grouping on.  Total pages: 774025
(d4) [2017-05-08 02:14:36.422] [    0.000000] Policy zone: Normal
(d4) [2017-05-08 02:14:36.422] [    0.000000] Kernel command line:
debug earlyprintk=xen root=/dev/xvda2 iommu=soft
(d4) [2017-05-08 02:14:36.423] [    0.000000] PID hash table entries:
4096 (order: 3, 32768 bytes)
(d4) [2017-05-08 02:14:36.566] [    0.000000] software IO TLB [mem
0x13de00000-0x141e00000] (64MB) mapped at
[ffff88013de00000-ffff880141dfffff]
(d4) [2017-05-08 02:14:36.578] [    0.000000] Memory:
2999308K/3145340K available (7868K kernel code, 1384K rwdata, 3560K
rodata, 1504K init, 704K bss, 146032
(d4) [2017-05-08 02:14:36.578] K reserved, 0K cma-reserved)
(d4) [2017-05-08 02:14:36.578] [    0.000000] Hierarchical RCU implementation.
(d4) [2017-05-08 02:14:36.578] [    0.000000]     Build-time
adjustment of leaf fanout to 64.
(d4) [2017-05-08 02:14:36.578] [    0.000000]     RCU restricting CPUs
from NR_CPUS=512 to nr_cpu_ids=2.
(d4) [2017-05-08 02:14:36.578] [    0.000000] RCU: Adjusting geometry
for rcu_fanout_leaf=64, nr_cpu_ids=2
(d4) [2017-05-08 02:14:36.580] [    0.000000] Using NULL legacy PIC
(d4) [2017-05-08 02:14:36.580] [    0.000000] NR_IRQS:33024 nr_irqs:48 0
(d4) [2017-05-08 02:14:36.580] [    0.000000] xen:events: Using FIFO-based ABI
(d4) [2017-05-08 02:14:36.581] [    0.000000] Console: colour dummy device 80x25
(d4) [2017-05-08 02:14:36.581] [    0.000000] console [tty0] enabled
(d4) [2017-05-08 02:14:36.581] [    0.000000] console [hvc0] enabled
(d4) [2017-05-08 02:14:36.581] [    0.000000] bootconsole [xenboot0] disabled


Guest's config:

name = "vm12"
kernel = "/home/kernels/vmlinuz-4.9.0-1-amd64"
initrd = "/home/kernels/initrd.img-4.9.0-1-amd64"
extra = "debug earlyprintk=xen root=/dev/xvda2 iommu=soft"
memory = 3072
vcpus = 2
cpus="12-13"
vif = [ 'mac=00:16:3e:38:3c:12' ]
disk = [ 
'phy:/dev/disk/by-path/ip-192.168.0.2:3260-iscsi-iqn.2017-02.ru.iqint.macavity:linux-vm-lun-1,hda,rw']
on_poweroff='destroy'
on_reboot='restart'
on_crash='destroy'
localtime=1
pci=['83:00.0,msitranslate=0']
e820_host=1

XEN boot options:

conring_size=5120K loglvl=all guest_loglvl=all iommu=1,verbose
com2=115200,8n1 console=com2 console_timestamps=datems
cpufreq=xen:performance dom0_mem=24G dom0_max_vcpus=6 dom0_vcpus_pin
flask=disabled

Dom0 boot options:

root=/dev/mapper/vg0-root ro iommu=1 intel_iommu=on
xen-pciback.permissive xen-pciback.passthrough=1
xen-pciback.hide=(04:00.0)(04:00.1)(82:00.0)(82:00.1)(83:00.0)(00:1a.0)(00:1d.0)
ro nomodeset debug earlyprintk=xen console=hvc0


Passed-through device in lspci:

...
82:00.0 VGA compatible controller: Advanced Micro Devices, Inc.
[AMD/ATI] Hawaii XT [Radeon R9 290X]
82:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aac8
83:00.0 RAID bus controller: Adaptec Series 8 12G SAS/PCIe 3 (rev 01)
ff:08.0 System peripheral: Intel Corporation Xeon E7 v2/Xeon E5
v2/Core i7 QPI Link 0 (rev 04)
ff:09.0 System peripheral: Intel Corporation Xeon E7 v2/Xeon E5
v2/Core i7 QPI Link 1 (rev 04)
...

Results of "xl pci-assignable-list" before starting domain:

# xl pci-assignable-list  | sort
0000:00:1a.0
0000:00:1d.0
0000:04:00.0
0000:04:00.1
0000:07:00.0
0000:82:00.0
0000:82:00.1
0000:83:00.0



Regards,
  Alexander

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
https://lists.xen.org/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.